Chromosome-level genome assembly of Agaricogonopus acrotrifoliolatus (Diplopoda: Spirostreptida) Authors: Chong Liu, JiaJia Wang, ShiJia Zhu, ZeWu Han, Xuankong Jiang, Yan Dong Published: 2025 ABSTRACT AbstractAgaricogonopus acrotrifoliolatus (Diplopoda: Spirostreptida) is the largest millipede species from China. Morphologically, it is characterized by a pair of prominent, fungiform gonopodal processes. Despite the ecological importance of Spirostreptida millipedes as key decomposers in soil ecosystems, genomic resources for this order are unavailable. Herein, we report the first chromosome-level genome assembly for Spirostreptida, represented by A. acrotrifoliolatus. The genome was constructed using a combination of PacBio HiFi long reads, Illumina short reads, and Hi-C sequencing. An initial assembly totaling 2.33 Gb was generated. Following Hi-C integration, 96.16% of the assembly was anchored to 11 pseudochromosomes, resulting in a final chromosome-level assembly of 2.24 Gb. The assembly exhibited high contiguity and completeness, with a scaffold N50 of 360.24 Mb and a 96.20% Benchmarking Universal Single-Copy Orthologs (BUSCO) score. Furthermore, 12,492 protein-coding genes and ~1.71 Gb (76.27%) of repetitive elements were identified. As the first reference genome for Spirostreptida, this assembly serves as a crucial genomic resource for advancing millipede genomics and evolutionary studies.Subject terms: Genomics, Genome, Eukaryote Background & Summary Millipedes (Diplopoda), as key decomposers, play vital roles in organic matter decomposition and nutrient cycling within soil ecosystems1,2. According to literature, the class Diplopoda comprises >12,000 described species across 145 families and 16 orders3,4. Most Diplopoda range 1–100 mm in body length5–13. In contrast, members of the order Spirostreptida are known for their exceptional size, often >300 mm in length5–13. This large body size enables Spirostreptida to efficiently fragment coarse organic debris, promoting early-stage decomposition and microbial colonization4,14. These traits underscore their ecological importance as macrodetritivores in soil ecosystems15. Despite their ecological relevance, genomic studies on Spirostreptida are unavailable, a gap also evident in public databases. As of July 2025, only 36 genome assemblies for Diplopoda were available in the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov/datasets/genome/?taxon=7553), representing 7/16 recognized orders: Chordeumatida, Glomerida, Julida, Platydesmida, Polydesmida, Polyzoniida, and Spirobolida. Among these, only three species have chromosome-level genome assemblies, including Cylindroiulus punctatus (GCA_965125795.1; 354.1 Mb, 7 chromosomes, scaffold N50 = 47,054.77 kb, order Julida), Choneiulus palmatus (GCA_965278725.1; 626.5 Mb, 14 chromosomes, scaffold N50 = 42,358.63 kb, order Julida), and Nanogona polydesmoides16 (GCA_965153395.1; 406.3 Mb, 16 chromosomes, scaffold N50 = 25,383.15 kb, order Chordeumatida). To date, no genomic resource of any kind has been reported for Spirostreptida. Agaricogonopus acrotrifoliolatus, a member of Spirostreptida, was first described from the tropical rainforests of Xishuangbanna, Yunnan, China17. It is characterized by a pair of prominent fungiform processes arising from the gonopods and a tripartite basal plate (gonocoxite)17. As the largest millipede species documented in China, it is morphologically distinctive and taxonomically representative. These attributes make A. acrotrifoliolatus an excellent candidate for genome sequencing to address the current lack of genomic data for Spirostreptida. In this study, we report the first chromosome-level genome assembly for the order Spirostreptida, represented by A. acrotrifoliolatus. The assembled genome spans ~2.24 Gb and includes 11 chromosome-scale scaffolds, with a scaffold N50 of 360.24 Mb. Individual chromosomes range in length from 53 Mb to 394 Mb. Benchmarking Universal Single-Copy Orthologs (BUSCO) analysis, based on the arthropoda_odb10 lineage dataset, revealed a high completeness score of 96.20%. The combination of a high scaffold N50 and strong BUSCO results indicates that the assembly is both highly contiguous and complete. As the first genome assembly for Spirostreptida, the genome of A. acrotrifoliolatus not only addresses a major gap in genomic data for this order but also provides a valuable foundation for future evolutionary and functional studies. Methods Sample collection and sequencing A male specimen of A. acrotrifoliolatus was collected from the Xishuangbanna Tropical Botanical Garden in Yunnan Province, China. Immediately after collection, the sample was preserved at −80 °C to maintain DNA integrity for genome sequencing. To minimize potential contamination from gut microbiota, we dissected the head, thorax, gonopods, and body muscle tissues separately and extracted genomic DNA from these tissues for downstream sequencing. A multi-platform sequencing approach was used to generate a chromosome-level reference genome, incorporating PacBio HiFi long-read sequencing, Illumina short-read sequencing, and Hi-C chromatin conformation capture technologies. For HiFi sequencing, genomic DNA was extracted using the Blood & Cell Culture DNA Midi Kit (Qiagen, Hilden, Germany). Following circular DNA template preparation, DNA fragments >20 kb were selected for SMRTbell library construction. Then, whole-genome sequencing was conducted on the PacBio Revio platform, yielding ~85 Gb of HiFi long-read data (Table 1). Table 1Statistics of the sequencing data.LibrariesSequencing platformraw dataclean dataDepth* (X)Pacbio HiFiPacBio Revio85,310,084,31385,310,084,313~37IlluminaIllumina HiSeq 2000117,771,967,200113,519,199,000~49Hi-CIllumina HiSeq 20007,970,022,9007,038,342,633—HiFi RNA-seqPacBio Revio4,322,382,6194,214,810,784—HiFi RNA-seqPacBio Revio2,170,671,8562,125,263,958—Illumina RNA-seqIllumina HiSeq 20007,496,587,5007,274,303,739—*For ease of calculation, the genome size of Agaricogonopus acrotrifoliolatus is assumed to be 2.3 Gb. Statistics of the sequencing data. *For ease of calculation, the genome size of Agaricogonopus acrotrifoliolatus is assumed to be 2.3 Gb. For Hi-C sequencing, a high-throughput chromatin conformation capture (Hi-C) library was prepared using formaldehyde crosslinking, DpnII restriction enzyme digestion, biotin labeling, proximity ligation, and purification18. The resulting libraries, with insert sizes ranging 300–600 bp, were sequenced on an Illumina HiSeq 2000 platform, yielding ~8 Gb of 150-bp paired-end reads (Table 1). For transcriptome sequencing. A male specimen of A. acrotrifoliolatus was dissected to collect head, legs, gonopods, and body muscle tissues along the anterior–middle–posterior gradient. Total RNA was extracted from these tissues using TRIzol reagent. mRNA was reverse-transcribed to cDNA, and RNA-seq libraries were constructed using the TruSeq RNA v2 kit, while full-length transcriptome libraries were prepared with the SMRTbell Express Template Prep Kit 2.0. Sequencing was performed on Illumina HiSeq 2000 and PacBio Revio platforms, generating both short-read and full-length transcriptome data, which were subsequently quality-controlled and used for genome annotation. For the genome survey, ~118 Gb of raw data were generated through paired-end sequencing (2 × 150 bp) on the Illumina HiSeq 2000 platform (Table 1). Genome survey The raw Illumina reads were processed using fastp (v0.23.4)19 to obtain high-quality clean reads for downstream analyses. Filtering criteria included (1) reads containing >5 bp of adapter sequence; (2) reads with low sequencing quality (≥50% of bases with Q-scores <19); and (3) reads with high ambiguity (N bases >5%). If either mate in a read pair failed to meet the quality criteria, the entire pair was discarded. Following quality filtering, k-mer analysis (k = 21) was performed using Jellyfish (v2.3.1)20, and genome features were inferred using GenomeScope2.021. The genome survey estimated a genome size of ~2.28 Gb, with a homozygosity rate of ~0.22%, a total repeat length of ~1.46 Gb, and a model fit of 97.1% (Fig. 1a). (a) K-mer frequency and genome size estimation of the Agaricogonopus acrotrifoliolatus genome. (b) Male specimen of A. acrotrifoliolatus.Scale bars: 10 mm. Genome assembly Prior to the initial genome assembly, raw HiFi long reads were screened for adapter contamination using HiFiAdapterFilt (v2.0.0)22 with default settings. No adapter-contaminated reads were detected, and the clean reads were retained in FASTQ format for subsequent assembly. The filtered HiFi long reads were assembled using Hifiasm (v0.19.5-r587)23 to generate the initial draft genome. This assembly spanned ~2.33 Gb and consisted of 1,276 contigs, with a contig N50 of ~31 Mb and a GC content of 41.51%. Assembly completeness was evaluated using BUSCO (v5.4.7)24 with the arthropoda_odb10 database, which identified 96.30% of BUSCO genes as complete (Table 2; Fig. 2). To anchor the genome assembly into chromosome-scale scaffolds, Hi-C analysis was performed. Initially, the raw Hi-C reads were quality-filtered using TrimGalore (v0.6.10) (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) with the following parameters: “trim_galore -q 20 --phred33 --fastqc --stringency 1 --length 20 -e 0.1 --paired --dont_gzip --clip_R1 4 --clip_R2 4 --three_prime_clip_R1 4 --three_prime_clip_R2 4 --illumina -o ./.” Using the cleaned Hi-C data, the draft genome was scaffolded to the chromosome level using the AutoHiC (v1.3.3)25 automated pipeline. This pipeline integrates genome indexing, the Juicer26 workflow, the 3D-DNA27 scaffolding protocol, and deep learning–based error correction and structural optimization. In total, 96.16% of the draft genome was anchored to 11 pseudochromosomes, resulting in a final chromosome-level assembly of ~2.24 Gb, with a scaffold N50 of 360.24 Mb (Table 2) and individual chromosome lengths ranging 53–394 Mb (Table 3). Chromosome numbering follows the default AutoHiC output order and does not carry biological significance. The completeness of the final genome assembly was assessed using BUSCO analysis, which identified 96.20% complete, 2.4% fragmented, and 1.4% missing BUSCOs (Fig. 2). These metrics indicate that the anchored genome is highly complete and of sufficiently high quality for downstream genomic and evolutionary analyses. Table 2Genome assembly statistics for Agaricogonopus acrotrifoliolatus.TypePreliminary AssemblyChromosome-level AssemblyAssembly size (bp)2,330,714,2042,241,317,908Contig/Scaffold N50 (bp)31,442,345360,240,644Contig/Scaffold N90 (bp)2,503,738123,269,442CC ratio1161Number of contig/pseudochromosome1,27611Longest contig/scaffold (bp)195,835,358393,078,066Contig/Scaffold L50183Contig/Scaffold L901039GC content (%)41.5141.37BUSCO completeness (%)96.3096.20 Genome assembly statistics for Agaricogonopus acrotrifoliolatus. BUSCO summary of the Agaricogonopus acrotrifoliolatus genome. The x-axis represents the percentage of BUSCOs; the y-axis indicates different genome assembly stages. Table 3Statistical summary of the 11 pseudochromosomes in the Agaricogonopus acrotrifoliolatus genome.ChromosomeLength(bp)GC(%)% of assemblyChromosome 1388,542,41041.8316.67Chromosome 253,865,10440.082.31Chromosome 3128,125,11039.845.5Chromosome 4277,036,02441.6911.89Chromosome 5123,269,44240.025.29Chromosome 6149,133,42640.296.4Chromosome 756,380,13640.652.42Chromosome 8126,334,15642.465.42Chromosome 9360,240,64441.4915.46Chromosome 10185,313,39040.187.95Chromosome 11393,078,06642.3816.87Total2,241,317,90841.3796.17 Statistical summary of the 11 pseudochromosomes in the Agaricogonopus acrotrifoliolatus genome. Genome repeats annotation A de novo repeat library specific to A. acrotrifoliolatus was constructed using RepeatModeler (v2.0.5)28. This custom library was integrated with the Dfam3.729 and RepBase-2018102630 databases to generate the final repeat sequence reference. Repetitive elements in the genome were annotated using RepeatMasker (v4.1.5)31 based on this combined library. The analysis revealed that ~76.27% of the genome consists of repetitive sequences, including long interspersed nuclear elements (LINEs, 34.10%), short interspersed nuclear elements (SINEs, 0.37%), DNA transposons (8.20%), long terminal repeat elements (LTRs, 7.58%), unclassified elements (25.40%), and other repeat types (Table 4; Fig. 3). Table 4Statistics of repeat elements identified in the Agaricogonopus acrotrifoliolatus genome.TypeNumber of elementsLength (bp)% of genomeLINE743,929764,347,60034.10SINE58,9658,240,4540.37LTR120,064169,814,3547.58DNA Transposon590,883183,792,9748.20Unknown1,579,163569,189,01825.40Other31,64714,055,3540.62Total3,124,6511,709,439,75476.27 Statistics of repeat elements identified in the Agaricogonopus acrotrifoliolatus genome. Genomic features of the assembled Agaricogonopus acrotrifoliolatus genome: (A) chromosomes, (B) repetitive sequence density, (C) GC content, (D) gene density, and (E) colinear relationships among chromosomes. Genome annotation Protein-coding genes were predicted following repeat masking, using three complementary strategies: RNA-seq-based, homology-based, and ab initio gene prediction. For transcriptome-based gene prediction, Illumina RNA-seq reads were aligned to the genome using HISAT2 (v2.1.1)32, and transcripts were assembled with StringTie (v2.1.0)33. Full-length transcriptomic reads from PacBio HiFi sequencing were filtered and clustered using IsoSeq. 3 (v4.0.0)34, and the polished HiFi isoforms were aligned to the genome using Minimap2 (v2.26)35. Both Illumina RNA-seq and PacBio full-length transcriptome data were subsequently used as evidence for transcriptome-based gene prediction. For homology-based annotation, protein sequences from five representative Diplopoda species, Anaulaciulus tonginus, Glomeris maerens, Helicorthomorpha holstii, Niponia nodulosa, and Trigoniulus corallinus, were retrieved from Figshare36 (10.6084/m9.figshare.15088722.v4). The final protein-coding genes were predicted using BRAKER3 (v3.0.3)37 in ETP mode, which integrates RNA-seq data, homologous proteins, and ab initio predictions to produce evidence-supported, high-confidence gene models. In total, 12,492 protein-coding genes were predicted. The average gene length was 26,861 bp. The BUSCO analysis indicated high annotation completeness, with 95.8% complete, 0.8% fragmented, and 3.5% missing BUSCOs. To functionally annotate the predicted protein-coding genes, the resulting protein sequences were aligned against multiple public databases, including eggNOG38. KOG39, GO40, KEGG41, and Pfam42, using BLASTP43. Consequently, 10,757 genes were functionally annotated in eggNOG, 10,356 in KOG, 8,797 in GO, 7,823 in KEGG, and 10,202 in Pfam (Table 5). These findings indicate that the majority of predicted genes have robust functional annotations across multiple databases, reflecting the gene set’s high quality and completeness. The annotation of noncoding RNAs was conducted independently. Transfer RNA (tRNA) genes were predicted using tRNAscan-SE (v2.0)44. and ribosomal RNA (rRNA) genes were identified using RNAmmer (v1.2)45 under the eukaryotic model. MicroRNAs (miRNAs), small nuclear RNAs (snRNAs), and other noncoding RNAs (ncRNAs) were annotated using INFERNAL (v1.1.2)46 via its cmscan program, based on the Rfam database (v14.6)47. In total, four major classes of noncoding RNAs were identified: 9,740 rRNAs, 1,046 snRNAs, 51 miRNAs, and 16,342 tRNAs (Table 6). Table 5Statistics of functional gene annotations in the Agaricogonopus acrotrifoliolatus genome.TypeTotalEggNogKOGGOKEGGPfamNumber12,49210,75710,3568,7977,82310,202Percentage100%86.11%82.90%70.42%62.62%81.67% Statistics of functional gene annotations in the Agaricogonopus acrotrifoliolatus genome. Table 6Statistics of noncoding RNA annotations in the Agaricogonopus acrotrifoliolatus genome.TypeNumberAverage Length (bp)Totoal Length (bp)miRNA5179.474,053tRNA16,34275.531,234,293rRNA18S2,9901,746.615,222,35328S3,5483,781.9913,418,5285S372118.7344,1685.8S2,830153.87435,458snRNAU1_snRNA290161.7546,908U2_snRNA120183.9822,078U5_snRNA70123.438,640U6_snRNA22107.732,370U4_snRNA11131.641,448U12_snRNA1153.00153Other_snRNA532158.4384,286 Statistics of noncoding RNA annotations in the Agaricogonopus acrotrifoliolatus genome. Data Records All sequencing data reported in this paper, including Illumina, PacBio HiFi, Hi-C, RNA-seq, and PacBio IsoSeq data, have been deposited in the Genome Sequence Archive (GSA) under accession number CRA02890148. The assembled genome has been deposited at GenBank under the accession number GCA_052040765.149. Technical Validation The final genome assembly spans 2.24 Gb, with a scaffold N50 of 360.24 Mb. This size is consistent with the genome size estimated through k-mer analysis and the genome survey (Fig. 1a). Assessment using BUSCO, based on the arthropoda_odb10 database, identified 96.2% of universally conserved genes, indicating that the A. acrotrifoliolatus genome assembly is nearly complete. Similarly, evaluation of the protein-coding gene set using BUSCO in protein mode (-m protein) revealed a completeness score of 95.8%. Acknowledgements This work was supported by National Natural Science Foundation of China (31401971), The Outstanding Youth Research Project of Colleges and Universities in Anhui Province (No. 2022AH030110), Scientific Research Starting Foundation of Chuzhou University China (2024qd42, 2025qd17). Author contributions Conceptualization: Yan Dong; Sample collection: JiaJia Wang, ShiJia Zhu, and ZeWu Han; DNA, RNA extraction and sequencing: JiaJia Wang, ShiJia Zhu, and ZeWu Han; Bioinformatics analysis: Chong Liu, JiaJia Wang and ShiJia Zhu; Visualization: Chong Liu and JiaJia Wang; Original draft preparation: Chong Liu; Manuscript review & editing: Chong Liu, JiaJia Wang, Xuankong Jiang and Yan Dong; Funding acquisition: Yan Dong; Final approval of the manuscript: All authors. Data availability All sequencing data reported are avaoilable at https://ngdc.cncb.ac.cn/gsa/browse/CRA028901. The assembled genome and annotations are available at https://identifiers.org/ncbi/insdc.gca:GCA_052040765.1. Code availability No custom scripts were developed in this study. All analyses were conducted according to the official manuals and standard protocols of the respective bioinformatics tools. Detailed software versions and parameters are provided in the Methods section. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Chong Liu, JiaJia Wang. Contributor Information Xuankong Jiang, Email: antoma93@gmail.com. Yan Dong, Email: dongyan_bio@126.com. References Associated Data Data Availability Statement All sequencing data reported are avaoilable at https://ngdc.cncb.ac.cn/gsa/browse/CRA028901. The assembled genome and annotations are available at https://identifiers.org/ncbi/insdc.gca:GCA_052040765.1. No custom scripts were developed in this study. All analyses were conducted according to the official manuals and standard protocols of the respective bioinformatics tools. Detailed software versions and parameters are provided in the Methods section.